Molnar et al. 2021: Forelimb function across the fish–tetrapod transition

Molnar et al. 2021
bring us a deep and complex look into the hypothetical muscles (based on muscle scars) of Eusthenopteron (Fig. 1), Acanthostega (Fig. 1) and Pederpes. The authors compare these distinctly different taxa to “show that early tetrapods share a suite of characters including restricted mobility in hurmerus long-axis rotation, increased muscular leverage of humeral retraction, but not depression/adduction, and increased mobility in elbow flexion-extension.” The authors infer the earliest ‘steps’ in tetrapod forelimb evolution were related to limb-substrate interactions. Weight support appeared later.

Figure x. The fin to finger transition in the LRT with the addition of Elpistostege.

Figure 1. The fin to finger transition in the LRT with the addition of Elpistostege.

Unfortunately, 
without a valid phylogenetic context, what these authors deliver is not quite germane to the topic of their headline. The actual fin-to-finger transition occurred between Panderichthys (Figs. 1, 2) and the extremely similar Trypanognathus (Figs. 1, 2). The former had fins. The latter had fingers and toes. Otherwise they were very much alike.

Molnar et al. looked at the wrong taxa. Neither Panderichthys nor Trypanognathus are mentioned in the Molnar et al. text.

What can we conclude given
the similarities and differences of Panderichthys and Trypanognathus?

  1. Small fins and limbs at the transition were incapable of weight bearing
  2. Elbows and knees were incapable of bending, pushing, pulling
  3. Torso much longer than tail, lots of flexible ribs
  4. Low, wide, flexible torso at the transition provided serpentine locomotion
  5. Little risk of tipping over due to low center of gravity
Figure 6. Dorsal and ventral views of Panderichthys and several basal tetrapods demonstrating the low, flat skulls and bodies with small limbs and relatively straight ribs.

Figure 2. Dorsal and ventral views of Panderichthys and several basal tetrapods demonstrating the low, flat skulls and bodies with small limbs and relatively straight ribs.

Molnar et al. conclude:
“Together, these results suggest that competing selective pressures for aquatic and terrestrial environments produced a unique, ancestral “early tetrapod” forelimb locomotor mode unlike that of any extant animal.”

Not really. Consider the moray eel chasing crabs on land without fins or fingers. Click the pic to view video on YouTube. David Attenborough is the narrator.

Now put predator and prey in a Devonian swamp setting,
with lots of growing and rotting vegetation and no rocky place to find safety. Note (in figure 2), the rather slow phylogenetic growth of the limbs relative to the torso in this sequence. Other lineages did their own thing in their own time. Ossinodus, for instance (Fig. 1), had a shorter torso and longer limbs, and was a phylogenetic ancestor to Ichthyostega and Acanthostega.


References
Molnar JL, Hutchinson JR, Diogo R, Clack JA and Pierce SE. 2021. Evolution of forelimb musculoskeletal function across the fish-to-tetrapod transition. Science Advances 2021; 7: eabd7457 22 January 2021

Where do Porolepiformes fit into the LRT?

Traditioinally there are four types of lobefin fish:

  1. Actinista = coelocanths (e.g. Latimeria, Fig. 1)
  2. Dipnoi = lungfish (e.g. Polypterus, Fig. 1)
  3. Osteolepiformes = stem tetrapods (e.g. Osteolepis, Fig. 1)
  4. Porolepiformes = all extinct (e.g. Porolepis, Fig. 1)
Figure 1. Lobefin fish clades in this subset of the LRT.

Figure 1. Lobefin fish clades in this subset of the LRT.

A fifth clade of lobefin fish
was recovered by the large reptile tree (LRT, 1787+ taxa, subset Fig. 1). The Stensioella + Guiyu clade does not nest with placoderms, as other workers posit. Those workers consider Stensioella ‘enigmatic’ and ‘with arcane affinity’. In the LRT Stensioella does not nest with coelochanthiformes either, but with Youngolepis, Guiyu and other flattened big-flipper fish, most of which do not preserve post-crania.

Getting back to Porolepiformes…
This clade starts with a Carboniferous late survivor of a Silurian radiation, Allenypterus (Fig. 2). This odd sort of ‘traditional coelocanth’ had a straight, eel-like tail, a tall, narrow torso and lobe pectoral fins only.

Figure 2. Allenypterus nests with the coelacanth lobefins in the LRT and elsewhere.

Figure 2. Allenypterus nests with the coelacanth lobefins in the LRT and elsewhere.

Quebecius
(Fig. 3) had lobe-fin pectoral fins, but all other fins were ray fins. The body was rounder in cross-section and the eyes were relatively smaller. Marginal teeth were tiny on longer jaws.

Figure 2. Quebecius is similar in most respects to Diplacanthus. The pectoral fin still has a pointed appearance, but the other fins have more typical rays.

Figure 3. Quebecius is similar in most respects to Diplacanthus. The pectoral fin still has a pointed appearance, but the other fins have more typical rays.

Holoptychius serrulatus
(Agassiz 1839, Cope 1897; Middle Devonian to Carboniferous, 50cm to 2.5m long, Fig. 4) had lobe-fins only. The anterior skull + body was rounder in cross-section and the eyes were relatively smaller. This genus is found in marine sandstone. Tiny teeth lined the jaws, but are rarely illustrated due to their minuscule size. Large fangs descend from the palate and rise from the coronoids.

Figure 2. Holoptychius is a basal lobefin in the coelacanth clade.

Figure 4. Holoptychius is a basal lobefin in the coelacanth clade.

Porolepis posnaniensis
(Kade 1858, named by Woodward 1891; Early Devonian, Fig. 5) apparently had a post-crania similar to Holoptychius. They eyes continued their decline in size. Colors are tetrapod homologs often different than labels. The upper squamosal is here the postorbital.

Figure 2. Holoptychius and Porolepis skulls compared.

Figure 5. Holoptychius and Porolepis skulls compared.

Laccognathus panderi 
(Gross 1941,Vorobyeva 2006, Down et al. 2011; Middle to Late Devonian, 390-360mya, Figs. 5, 6) was a marine costal or lagoon bottom dweller with an even wider skull, even smaller eyes and large palatal and coronoid fangs. The two external nares are confluent. Post-crania is largely unknown.

Figure 3. Laccognathus diagram from Downs et al. 2011. Colors and tetrapod homolog labels added.

Figure 6. Laccognathus diagram from Downs et al. 2011. Colors and tetrapod homolog labels added.

Figure 5. Laccognathus specimen in situ from Downs et al. 2011. Colors added.

Figure 7. Laccognathus specimen in situ from Downs et al. 2011. Colors added.

According to Wikipedia
“Porolepiformes was established by the Swedish paleontologist Erik Jarvik (1980), and were thought to have given rise to the salamanders and caecilians independently of the other tetrapods. He based this conclusion on the shapes of the snouts of the aforementioned groups. This view is no longer in favour in Paleontology (Schultz and Trueb 1991).”

“Jarvik also claimed the existence of choanae in porolepiformes which linked them to tetrapods, but this has remained controversial. (Clement 2001). Recent phylogenetic reconstruction places porolepiformes close to lungfishes (Janvier 1996).”

In the LRT 
porolepiformes are closer to coelacanths, catfish and placoderms. Lungfish are closer to tetrapods. Choanae are not present in porolepiformes (Clement 2001).


References
Ahlberg PE1991. A re-examination of sarcopterygian interrelationships, with special reference to the Porolepiformes.
–Zoological Journal of the Linnean Society: Vol. 103, #3, pp. 241-287 [doi: 10.1111/j.1096-3642.1991.tb00905.x]
Clement G 2001. Evidence for lack of choanae in the Porolepiformes. Journal of Vertebrate Paleontology, 21: 795–802.
Downs J, Daeschler E, Jenkins F Jr and Shubin N 2011. A New Species of Laccognathus(Sarcopterygii, Porolepiformes) from the Late Devonian of Ellesmere Island, Nunavut, Canada. Journal of Vertebrate Paleontology. 31 (5): 981–996.|
Janvier P 1996. Early vertebrates. Oxford science publications. 1996, Oxford, New York: Clarendon Press; Oxford University Press.|
Jarvik E 1980. Basic structure and evolution of vertebrates. Vol. 1-2. Academic Press (London).
Schultze H-P and Trueb L1991. Origins of the higher groups of tetrapods: controversy and consensus. Cornell University Press. p. 37.
Vorobyeva EI 2006. A new species of Laccognathus (Porolepiform Crossopterygii) from the Devonian of Latvia. Paleontol. J. Physorg.com. 40 (3): 312–322. doi:10.1134/S0031030106030129.
Woodward AS 1891. Catalogue of the Fossil Fishes in the British Museum (Natural History). Part II. Catalogue of the Fossil Fishes in the British Museum (Natural History) 2.

wiki/Porolepiformes
wiki/Porolepis
wiki/Holoptychus
wiki/Allenypterus
wiki/Quebecius
wiki/Actinistia
wiki/Coelocanth

The phylogenetic comings and goings of lobe fins

The distribution of lobe fins, spiny fins and ray fins 
in the large reptile tree (LRT, 1756+ taxa, subset Fig. 1) indicate that each type of fin came and went and sometimes came back again.

So, contra traditional paleontology,
each type of fin does not represent a monophyletic clade. That would be “Pulling a Larry Martin” by setting up clades based on just a few characters.

The LRT provides a more holistic approach,
looking at 238 character traits from nose to toes and letting the software decide without tradition or bias. The LRT documents the multiple evolution of ray fins by convergence.

Figure 1. Subset of the LRT focusing on basal vertebrates and highlighting ray fins, spiny sharks and lobe fins. Catfish retain spines on their ray-like pectoral fins.

Figure 1. Subset of the LRT focusing on basal vertebrates and highlighting ray fins, spiny sharks and lobe fins. Catfish retain spines on their ray-like pectoral fins.

The basalmost lobefin in the LRT has gone unrecognized until now,
perhaps because Ticinolepis longaeva (Fig. 2) has such a little lobe on its pectoral fin and it is only known from Middle Triassic fossils. Ticinolepis longaeva nests at the base of all lobefins in the LRT (subset Fig. 1) so it would have had a Silurian genesis.

Figure 2. Ticinolepis longaeva in situ and reconstructed. Note the pectoral fin has a small lobe and this taxon nests at the base of all lobefins in the LRT.

Figure 2. Ticinolepis longaeva in situ and reconstructed. Note the pectoral fin has a small lobe and this taxon nests at the base of all lobefins in the LRT.

The resemblance of Ticinolepis longaeva to the next most basal lobefin,
Miguashaia (Middle Devonian; Fig. 3) is also instructive. (As a side note, Ticinolepis crassidens nests with Perleidus, not with Ticinolepis longaeva in the LRT, contra López-Arbarello and Sferco 2018).

Figure 2. The lobefin Miguashaia. Compare to the spiny shark, Diplacanthus, figure 1.

Figure 3. The lobefin Miguashaia. Compare to the spiny shark, Diplacanthus, figure 1.

Ticinolepis longaeva
(López-Arbarello and Sferco 2018; 12cm; Middle Triassic; MCSN 8072) nests at the base of the lobefin fishes. Note the tiny lobe in the middle of the ray fin. Compare that pectoral fin to the one in figure 3.

Miguashaia bureaui
(Schultze 1973, Cloutier 1996; Middle Devonian; 45cm) was considered the sister group (outgroup) of the Actinista (coelocanths). Notably Miguashaia reverses to a heterocercal tail. That’s why it looks a little odd. The dentary is short and the teeth are small.

Figure 2. Sturgeon swimming in a test tank from Wilga and Lauder 1999.

Figure 4. Sturgeon swimming in a test tank from Wilga and Lauder 1999.

Final notes to be covered in more detail later:
Basal pectoral fins are rather inflexible and extend horizontally (Fig. 4). Ratfish hold their pectoral fins vertically, against the torso. Iniopterygians raise the pectoral fin to the dorsal margin. Moray eels lose their fins. So there is more variety here yet to explore.


References
López-Arbarello A and Sferco E 2018. Neopterygian phylogeny: the merger assay. Royal Society open sci. 5: 172337. http://dx.doi.org/10.1098/rsos.172337
Schultze H-P 1973.
 Crossopterygier mi heterozerker Schwanzfloss aus dem Oberdevon Kanadas, nebst einer Beschreibung von Onychodontida-Resten aus dem Middledevon Spaniens und aus dem Karbon der USA. Palaeontograhica A 143:188–208.

the large reptile tree

wiki/Miguashaia

Ticinolepis: DGS rebuilds scattered skull parts

Updated November 06, 2020
with new reconstructions of the skulls of Ticinolepis crassidens and Ticinolepis longavea. When both taxa were added to the LRT they did not nest with each other.

López-Arbarello, et al. 2016
and López-Arbarello and Sferco 2018 brought us two ostensibly congeneric Middle Triassic ganoid-scaled fish. These two species of Ticinolepis (Fig. 1) are distinguished by their teeth (and other traits). The larger one (T. longaeva, MCSN 8072) has small, slender teeth (Fig. 2). The smaller one (T. crassidens, PIMUZ T 273) has large bulbous teeth. Both were added to the large reptile tree (LRT, 1756+ taxa) and they did not nest together.

Figure 1. Two new Ticinolepis species to scale.

Figure 1. Two new Ticinolepis species to scale from López-Arbarello 2016, Scale bar = 2 cm. The skull of the top species, T. longavea, is shown in figure 2.

Ticinolepis crassidens
(López-Arbarello and Sferco 2016; Middle Triassic; PIMUZ T 273) nests with Perleidus in the LRT.

Figure 2. The skull of Ticinolepis crassidens (MCSN 8072) in situ from López-Arbarello 2016, traced and reconstructed using DGS methods.

Figure 2. The skull of Ticinolepis crassidens (MCSN 8072) in situ from López-Arbarello 2016, traced and reconstructed using DGS methods.

Ticinolepis longaeva
(López-Arbarello and Sferco 2016; Middle Triassic; MCSN 8072) nests at the base of all lobefin fish in the LRT, close to Mis

Figure 2. Ticinolepis longaeva in situ and reconstructed. Note the pectoral fin has a small lobe and this taxon nests at the base of all lobefins in the LRT.

Figure 2. Ticinolepis longaeva in situ and reconstructed. Note the pectoral fin has a small lobe and this taxon nests at the base of all lobefins in the LRT.

López-Arbarello and Sferco 2018 wrote:
“In our analysis, †Ticinolepis also joins the tree at this stem as the most basal Ginglymodi and, thus, we now find it useful to distinguish the clade (Lepisosteiformes, †Semionotiformes) as the Neoginglymodi, defined as the clade including Lepisosteus and †Semionotus, and all descendants of theirmost recent common ancestor.”

Currently in the LRT, that clade includes just those two sister taxa. Over time perhaps others will be added.

López-Arbarello and Sferco 2018 mention all LRT clade members, except Tarrasius. The authors do not attempt a reconstruction of either Ticinolepis. 


References
López-Arbarello A, Burgin T, Furrer H and Stockar R 2016. New holostean fishes (Actinopterygii: Neopterygii) from the Middle Triassic of the Monte San Giorgio (Canton Ticino, Switzerland). Peerj 4, 61 (doi:10.7717/peerj.2234).
López-Arbarello A and Sferco E 2018. Neopterygian phylogeny: the merger assay. Royal Society open sci. 5: 172337. http://dx.doi.org/10.1098/rsos.172337

Spiny sharks (Acanthodii) transitional to lobefins in the LRT

The most recent changes
to the large reptile tree (LRT, 1583 taxa, subset Fig. 1) resolve earlier problems and place two spiny sharks (clade: Acanthodii, Fig. 1) at the base of the newly expanded pre-lobefin clade, all arising from catfish + placoderms, some of which also have spiny pectoral fins.

Figure 1. Classic reconstruction of Cladoselache, a shark-like taxon basal to sturgeons and catfish+placoderms in the LRT.

Figure 1. Classic reconstruction of Cladoselache, a shark-like taxon. Note the robust pectoral fin skeleton. A Silurian sister is the genesis for the spiny sharks, catfish and placoderms.

These in turn
arise from taxa like Cladoselache (Fig. 1), which had strongly supported pectoral fins along with robust anterior spines on the two dorsal fins, homologs of dorsal spines on spiny sharks.

Figure 2. Updated subset of the LRT focusing on basal vertebrates (fish). Arrow points to Hybodus. This tree does not agree with previous fish tree topologies.

Figure 2. Updated subset of the LRT focusing on basal vertebrates (fish). Arrow points to Hybodus. This tree does not agree with previous fish tree topologies.

The tiny size of the basal acanthodian, Brachyacanthus, 
(Fig. 3) documents phylogenetic miniaturization at the genesis of a new major clade. If large eyes, a high forehead and a short rostrum indicate ‘cuteness’ and neotony, then Brachyacanthus is an early example of this. Cladoselache (Fig. 1) has two out of three of these traits.

According to Wikipedia
“Acanthodii or acanthodians (sometimes called spiny sharks) is a paraphyletic class of teleostomefish, sharing features with both bony fish and cartilaginous fish. In form they resembled sharks, but their epidermis was covered with tiny rhomboid platelets like the scales of holosteans (gars, bowfins). They represent several independent phylogenetic branches of fishes leading to the still extant Chondrichthyes.” In the LRT spiny sharks don’t lead to sharks, rays and chimaera, but diverge away from them.

Figure 2. The placoderm/catfish to spiny shark/lobe fin transition. We need more taxa, but here's how the LRT recovers it.

Figure 3. The placoderm/catfish to spiny shark/lobe fin transition. We need more Silurian taxa, but here’s how the LRT recovers it. Brachyacanthus, once again, documents phylogenetic miniaturization at the genesis of new major clades.

So, what is it about the spine fin
that made it a key trait?

Figure 1a. Cheirolepis fossils.

Figure 4. Cheirolepis fossils. Both have a spiny pectoral fin leading edge.

On the ray fin side of the cladogram
basal taxa include Pholidophorus (Fig. 5) and Coccocephalichthys (Fig. 6). This clade embraced open water speedy swimming and predation as their niche from the start. The extant tuna (Thunnus) is an extant relative of these two. Later taxa, like the frogfish (Antennarius) and sea robin (Prionotus), reverted to bottom-dwelling.

Figure 5. Pholidophorus ghosted to highlight the fins.

Figure 5. Pholidophorus fossil ghosted to highlight the fins and eyes.

By contrast, lobefins and their predecessors
appear to have preferred a slower swimming, bottom-dwelling lifestyle. That’s how they readily transitioned into shallow waters, swampy waters, swampy land and dry land in that order.

Figure 2. Coccocephalichthys (formerly Coccocephalus) is a Late Carboniferous transitional taxon between Devonian Strunius and Cretaceous Saurichthys.

Figure 6. Coccocephalichthys (formerly Coccocephalus) is a Late Carboniferous transitional taxon between Devonian Strunius and Cretaceous Saurichthys.

Even so,
some highly derived lobefins learned how to climb trees, fly, and even speed through open waters, with or without fins (Homo, Orcinus, Pavo).

The origin of the tetrapod quadratojugal

As we’ve seen
over the past several dozen fish additions to the large reptile tree (LRT, 1583 taxa) facial bones homologous with those of tetrapods often divide and fuse. We’ve already seen multipart jugals, lacrimals, nasals and squamosals in fish. This can be confusing and is probably the reason why fish facial bones are traditionally not labeled with tetrapod nomenclature. Expect homology arguments to last for decades, but here all fish facial bones are colored with tetrapod homologs.

Figure 1. Gogonasus skull demonstrating the genesis of the split between the toothy maxilla and the toothless quadratojugal.

Figure 1. Gogonasus skull demonstrating the genesis of the split between the toothy maxilla and the toothless quadratojugal.

The one bone that first appears by a split
of the maxilla into anterior toothy and posterior toothless portions is the quadratojugal. Phylogenetically the quadratojugal first appears on Gogonasus (Fig. 1). Prior taxa lack it. Later taxa have it. Even so, until the cladogram got figured out, this was puzzling.

Figure 1. Subset of the LRT alongside the definitions published in Laurin, Girondot and de Ricqles 2000.

Figure 2. Subset of the LRT alongside the definitions published in Laurin, Girondot and de Ricqles 2000. Note the Gogonasus node, where the quadratojugal first appears.

Gogonasus andrewsae (Long 1985, Long et al. 1997; Late Devonian, 380 mya; NMV P221807; 30-40cm in length) is the best preserved specimen of its type. This is the crossopterygian transitional between rhizodontids, coelocanths and higher tetrapodomorphs. The maxilla splits in two creating the tetrapod quadratojugal. The squamosal splits in two creating the tetrapodomorph preopercular. A pineal opening appears, so does the tetrapodomorph choana. The lacrimal contacts the external naris. This is the crossopterygian that gave rise to tetrapodomorphs and tetrapods, rhizodontids and gulper eels.


References
Long JA 1985. A new osteolepidid fish from the Upper Devonian Gogo Formation of Western Australia, Recs. Western Australia Mueum 12: 361–377.
Long JA et al. 1997. Osteology and functional morphology of the osteolepiform fish Gogonasus Long, 1985, from the Upper Devonian Gogo Formation, Western Australia. Recs. W. A. Mus. Suppl. 57, 1–89.

wiki/Gogonasus